40 MEMOIRS OF THE NATIONAL ACADEMY OE SCIENCES. 



it seems to be the type of a distinct snbdivisiou of the CeratocauipidiB (the subfamily Affliinw), as 

 we have iiaineil the group. This form is a connecting link between the genuine Ceratocampidie 

 and theSaturniid;e. As originally shown by Duponchel and more recently by Mr. Toulton, before 

 the last molt the caterpillar wholly discards the cougeuital characters, viz, its spinous armature, 

 characters common to the Ceratocampida-, and assumes an entirely and strikingly ditt'erent shape. 

 It is ni)w destitute of any spines at all, its body is rather short and thick, the segments full convex, 

 and it closely ap[)roxiuiates the geneval appearance of a Saturuiau larva, though the majority of 

 these are armed with more or less spinose tubercles; the caterpillar of the Brazilian Attucus bit is 

 Walker, however, is tigured and described by Burmeister as being smooth-bodied. The Aglia is 

 now quite a different creature from what it was in its earlier stages ; its conditions of existence have 

 somewhat changed, and in adaptation to such changes its means of defense are of a different nature. 

 It now feeds passively on its food plant, and is dependent on its colors, various delicate shades 

 of green and yellow, to escape the observation of its vertebrate and insect enemies, and if discovered 

 it appears to rely on its large, terrifying eye-spot and somewhat sphinx like attitude to frighten 

 away its aggressors. 



Now, it seems most natural to suppose that the features of the last stage were in the ancestors 

 of this insect acquired wholly or in part during a dehuite epoch in the lifetime of one or perhaps 

 of a few generations. The mature characters were not originally congenital and would have 

 perhaps been useless in the early stages of the caterpillar. They may have suddenly appeared in 

 a single individual and then have become transmitted by heredity and tixed by natural selection, or 

 this process may have extended through several, though not very many, generations. 



The chief factors in the origination of such a striking change iu shape and ornamentation after 

 the last molt appear to have been the atrophy of the spines and tubercles by disuse, the larva, by 

 a change in its mode of life, with more sluggish habits and perhaps feeding in less exposed 

 conditions, not needing them, the same change resulting in a transfer of the nutritive liuids 

 and bringing about the deposit of i)iginent in definite places, as in the eye-spots. 



Whether one accepts the view of the transmission of acciuired characters or not, it must be 

 conceded that the remarkable changes exhibited by Aglia in the last stage must have been 

 induced with more or less suddenness; that the tendency, at least, to the change was probably 

 originated during the lifetime of perhaps a single individual. The case seems to us to almost 

 amount to a crucial one, and if it can be explained by any other mode of reasoning than 

 the one suggested it will be a matter of interest. Certainly the congenital characters show a 

 remarkable contrast with what we assume to be acquired characters, and we know of no better 

 example which could be cited to prove the fact of the transmission of acquired characters. 



Family SaturniidK.—ln the larval stages of this regal family we have great contrasts between 

 the tirst and later stages, both in armature and coloration, as summarized below, each stage 

 differing remarkably from the others: 



RECAPITULATION OF THE MORE SALIENT ONTOGENETIC FEATURES OF PLATYSAMIA CECROPIA. 



A. COXGEXITAi CHAIiACTICUS. 



1. The setic in Stage I blunt, slightly bulbous, and glandular. 



2. The tubercles are all of tlie same size. 



3. Body in Stage I dark, almost blackish green; head jet black ; tubercles yellowish green. 



4. The homologue of the "caudal horn" shows plainly its double origin. 



5. The difference between the colors of the larva of the lirst and last stages very marked. 



li. EVOLUTION OF LATER ADAPTATIOXAL FEATIHES. 



1. The thoracic dorsal tubercles in Stage II and onward are longer than the abdominal ones. 



2. Five rows of indistinct black spots along the body in Stage II, not so distinct as in <S'. Cynthia^ 

 the body ])eing still dusky green. (These do not originate from lines.) At the end of Stage II 

 the larva is more like cynthia of the same age, tlie body being more yellow and the black spots 

 more distinct. The spots disappear at the end of Stage IV. 



